WebCam Face Recognition with data base of faces using JavaCV JavaFX GUI -- 2

Real time Face Recognition with data base of faces using JavaCV. General. The app will take a camera feed from webcam, usb, or wifi and recognise faces using EigenFaces from a data base created from training faces. A similar example can be found here - [login to view URL] [login to view URL] will be used for face detection. The project will be delivered as source, nbproject, and build. Specific. Important* - Using either openCV-2410 or OpenCV -300 and bytedeco JavaCV, all library jar files used in the project must be provided (to avoid current updates to v3 at bytedeco). The app should be set up for windowsx64 (opencv and javacv) but be clearly commented what library was used. The app will consist of a Main. gui 1024 x 768 horizontal with Camera feed to a top right panel 360 x 240. Camera feed to the panel can therefore be scaled down. A standard menu bar will be added but no functionality required. A text label underneath will display the name of the recognised face. The recognised face variable must be public static. The camera feed will initiatize with the main (no start button for camera) but should be running in its own thread/class. Sample scripts for Javacv are available on github at [login to view URL] Note: sample scripts at [login to view URL] will be read only from the 25th august. Example face detection available at [login to view URL]~ad/jg/nui07/ and [login to view URL] Although these will not work with opencv-300. The recogniser method takes cached data from [login to view URL] on run. A button from Main will open/set visible a training gui. Users can view their image 640x480 from camera feed with facial detection. The training gui will have four buttons, Save image, Load image from database, Train, Delete image from database and a progress bar on Train/close method. A textfield will collect the name of the user. Save the face button will save a .png image inside the detected triangle, grey scaling and clipping as necessary to produce an image approximately 15to20kb. The image is written into a folder getText() from the textField. If no text is found, a message dialog appears, else, save the file to /TrainingImages/. The Load image from database button calls jFileChooser to load .png from the TrainingImages/ directory and displays filename in textfield. If the the delete button is clicked, a confirmation box appears. On yes, the file is deleted. Both the Train button and close action calls the training script to run and train based on what is in the current TrainingImages folder - updating the eigen.cache. The progress bar measure progress on the training action. Before payment the delivered code will be tested and benchmarked for performance.